This topic is really only for publishing and logistics nerds. Since I am not nerdy enough, I am not really qualified to opine on the matter (but when did that stop anybody?). Anyway we find it an intriguing and perplexing issue. PersonaNonData today has a report on the flux that digital publishers find themselves in. Should there be as many ISBNs for each title as there are conceivable ebook formats? If so, there are going to be a very great many ISBNs, since it seems quite feasible that there are going to be a dozen, or perhaps many more. ebook/digital formats. Sure the market will settle down in due course to a few favoured formats. But that could take a while, and in the interim the ISBN system will need to cater for a very large set of potential numbers.
I have a suggestion: where titles go into a format where there are in effect many individual instances of the work then that format should have a separate ISBN attached to it. The ISBN system was introduced so that books would have a standard method of stock control. ISBNs are SKU’s. So digital platforms where copies of books are handed/downloaded to readers/purchasers the SKU specific to that channel serves a purpose. For digital platforms which are based on an ‘access’ system, which would include Google Book Search, and Amazon Search Inside, there is no need for a separate ISBN, because there are no ‘units’ that need to be tracked. Exact Editions is another such access system and there is no need therefore for publishers to assign separate ISBNs to their titles in the Exact Editions platform. The identifiers that matter for ‘access’ systems are the urls which comprise the book’s web presence.
I suspect that Exact Editions can hide behind the skirts of Google Book Search in this issue. It is pretty unlikely that Google will be prevailed upon to find and provide separate ISBNs for the millions of titles in its database. Very unlikely, because the ISBN fees for such a large number of titles will be a tidy sum. Very unlikely, because for many of the titles in Google Book Search, Google has no better idea than anybody else to whom the ISBN should be assigned. One of the difficulties with the Google Book Search project is that it is unclear who owns what. Who needs to be consulted about what? If Google knew how to assign ISBNs it would know which were the publishers to approach for permission to do so. Might as well ask them for permission to database the book at the same time?